An implant for osteosynthesis of the spine is already known, in particular from document FR-A-2 659 546, which implant comprises a pedicular screw provided with a head that is generally tuningfork-shaped, capable of directly receiving a cylindrical rod for linking vertebrae together. A split ring and a threaded clamping plug acting between the two branches of the head enable the assembly to be locked with angular adjustment being possible between the rod and the axis of the pedicular screw.
Installing that type of implant does not present the surgeon with any major difficulty, providing the heads of the various pedicular screws are sufficiently well aligned with one another, since under those circumstances the intervertebral link rod can be put into place relatively easily in said heads, without excessive stresses or twisting.
However, particularly depending on the type of pedicular aiming performed by the surgeon, and also depending on the orientation of the pedicles specific to each patient, it often happens that the axes of the pedicular screws are significantly inclined relative to the sagittal plane, and as a result, the housings for receiving the intervertebral link rod and as defined by the heads of the various screws can be very significantly out of alignment.
Under such circumstances, the solution which consists in deforming the rod very significantly so as to force it to follow the imposed path is either impracticable because of the considerable deformation forces that would be required, or else dangerous because it runs the risk of weakening the rod.
Various other devices are known for the purpose of making it possible to fix and lock a rod that is laterally offset relative to a pedicular screw.
Examples of such devices are given in the following documents: DE-A-195 12 709, U.S. Pat. Nos. 5,575,791, 5,002,542, WO-A-95 02372, and WO-A-96 29947.
Nevertheless, all of those known devices suffer from drawbacks. Thus, some of them are of considerable size, particularly in height, which is quite undesirable and considerably reduces the advantage in terms of compactness of "tuningfork"-type screw devices of the kind described above. Others are complex in structure, having numerous parts. Others are structurally incapable of receiving rods at an appropriate distance from the pedicles, which distance is moreover imposed by the "tuningfork"-type screws. Finally, some of those known devices are unsuitable for locking the rod relative to its anchor point in the bone in a manner that is sufficiently firm and strong.